Developing practical methods for extracting and analyzing objective behavioral, linguistic, and social- networking data from e-health interventions is crucial for identifying and evaluating potential mechanisms of action for these interventions. Understanding how users engage with intervention content, the social networks in which individual engagement is embedded, and the quality of the interactions that occur between users will be instrumental in improving existing e-health interventions and making them more effective. Only a handful of Internet-based interventions has been tested for cancer survivors, and most randomized trials have shown evidence of positive outcomes. Among the studies showing positive effects, a common treatment element is the use of social-networking features, such as professional facilitation, discussion boards, private messages, and other ways for survivors to make personal connections with similar others. This study will plug several large gaps among methodological tools available for analysis of Internet-based interventions. First, the study will use objective behavioral data to identify several longitudinal markers of individual engagement with an intervention, including time spent using the intervention, time spent in skills-training exercises, and word count. Second, the study will identify specific social-networking attributes and markers of the quality of support received that can be used to predict intervention engagement and outcomes over time. The aims of the study are: 1) to identify, characterize, and compare social-networking attributes of the STC and HSN interventions, 2) to evaluate the longitudinal effects of social-networking attributes and social interaction quality on exposure to the intervention (i.e., dose of treatment), and 3) to evaluate the effects of social-networking attributes and social interaction quality on outcomes of the interventions. Behavioral, linguistic, and self-report data will be derived from two of the largest, Internet-based survivorshp trials: Surviving and Thriving Cancer (STC, n = 352) and Health-Space.net (HSN, n = 200). Linguistic (i.e., text) data will be subjected to automated text analysis to generate markers of quality of social support. Patterns of interactions between participants will be used to generate actor-other matrices, which will be subjected to social network analysis. Website use data will be used to generate markers of individual engagement with each intervention. Statistical analyses will be used to evaluate the effects of social-networking attributes and interaction quality on engagement and outcomes across time. These results could be used to quickly identify subgroups at risk for low-engagement with an intervention, to tailor intervention content based on social network attributes or quality of support, or to benchmark the network properties of other group-based e-health interventions. Given the substantial reach available to Internet-based interventions, even relatively modest improvements to outcomes (e.g., by improving levels of engagement) would have the potential to greatly improve the public health impact of these kinds of interventions.